Alzheimer's diagnosis by an efficient pipelined gene selection model based on statistical and biological data...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Alzheimer's disease (AD) diagnosis using gene expression data from microarrays is an active research area.
• Microarray technology offers whole-genome data, potentially improving diagnostic accuracy.
• Platform-specific biases in microarray data analysis can negatively impact diagnostic performance.

Purpose of the Study:

• To develop a robust pipeline for diagnosing Alzheimer's disease.
• To integrate statistical analysis and artificial intelligence for improved diagnostic accuracy.
• To address data biases in microarray analysis for reliable AD diagnosis.

Main Methods:

• A pipeline approach combining statistical analysis and artificial intelligence techniques.
• Utilizing B-statistics to identify differentially expressed genes.
• Introducing an 'evidence score' based on gene interaction networks to assess biological relevance.
• Employing a genetic algorithm to select optimal gene subsets for classification.

Main Results:

• The proposed pipeline demonstrates effective identification of differentially expressed genes and their biological evidence.
• Artificial intelligence, specifically a genetic algorithm, successfully identified gene subsets for high sample separability.
• The method achieved fruitful predictive performance in diagnosing Alzheimer's disease, outperforming existing methods.

Conclusions:

• The developed pipeline offers a promising approach for accurate Alzheimer's disease diagnosis.
• Integrating statistical and AI methods effectively handles data biases and enhances biomarker discovery.
• The study provides a valuable tool for advancing Alzheimer's disease diagnostics through gene expression analysis.